Journal list menu
The Number of Point Mutations in Induced Pluripotent Stem Cells and Nuclear Transfer Embryonic Stem Cells Depends on the Method and Somatic Cell Type Used for Their Generation
Abstract
Induced pluripotent stem cells hold great promise for regenerative medicine but point mutations have been identified in these cells and have raised serious concerns about their safe use. We generated nuclear transfer embryonic stem cells (ntESCs) from both mouse embryonic fibroblasts (MEFs) and tail‐tip fibroblasts (TTFs) and by whole genome sequencing found fewer mutations compared with iPSCs generated by retroviral gene transduction. Furthermore, TTF‐derived ntESCs showed only a very small number of point mutations, approximately 80% less than the number observed in iPSCs generated using retrovirus. Base substitution profile analysis confirmed this greatly reduced number of point mutations. The point mutations in iPSCs are therefore not a Yamanaka factor‐specific phenomenon but are intrinsic to genome reprogramming. Moreover, the dramatic reduction in point mutations in ntESCs suggests that most are not essential for genome reprogramming. Our results suggest that it is feasible to reduce the point mutation frequency in iPSCs by optimizing various genome reprogramming conditions. We conducted whole genome sequencing of ntES cells derived from MEFs or TTFs. We thereby succeeded in establishing TTF‐derived ntES cell lines with far fewer point mutations. Base substitution profile analysis of these clones also indicated a reduced point mutation frequency, moving from a transversion‐predominance to a transition‐predominance. Stem Cells 2017;35:1189–1196
Significance Statement
iPS cells hold great promise for regenerative medicine but point mutations have been identified in these cells and have raised serious concerns about their safe use. We conducted whole genome sequencing of nuclear transfer ES cells derived from MEFs or tail‐tip fibroblasts (TTFs). We thereby succeeded in establishing ntES cell lines with far fewer point mutations. Our present results thus show that point mutations in iPS cells are therefore not a Yamanaka factor‐specific phenomenon but are intrinsic to genome reprogramming. Importantly, our results suggest that a greatly reduced point mutation frequency in induced pluripotent stem cells is feasible by optimizing various genome reprogramming conditions.
Introduction
Whereas induced pluripotent stem cells (iPSCs) hold great promise for use in regenerative medicine 1-8, the identification of a substantial number of point mutations in their genomes 9-15 has aroused concerns that they may not be safe to use, since such mutations could have tumorigenic or immunogenic effects 16-19. Therefore, in addition to the study of the molecular mechanisms underlying these genetic aberrations, the issue whether the point mutation frequency in iPSCs can be reduced has become a focus of attention. We have here studied another type of genome reprogrammed pluripotent stem cells, nuclear transfer embryonic stem cells (ntESCs), to address whether point mutations generally arise during genome reprogramming in these cells. Furthermore, we explored whether it would be possible to reduce the point mutation frequency during genome reprogramming.
Materials and Methods
Mouse Strains and Parent Somatic Cells
The cell lines used in this study were generated from the C57BL/6J mouse strain (Japan SLC, Hamamatsu, http://www.jslc.co.jp/english/index2.htm) except for a set of three ntES sister lines that were generated from a C57BL/6N mouse (Japan SLC). We used two mouse embryonic fibroblast (MEF) fractions and two tail‐tip fibroblast (TTF) fractions as the parent somatic cells, and each was prepared from an independent single embryo or adult. Animal experiments were approved by the Institutional Animal Care and Use Committees of the National institute of Radiological Sciences and University of Yamanashi.
Nuclear Transfer and Establishment of ntES Cell Lines
Nuclear transfer was performed as described previously 20. Cultured TTFs or MEFs were used as nuclear donors. After nuclear transfer, the reconstructed oocytes were activated using 5 mM SrCl2 in Ca‐free Chatot‐Ziomek‐Bavister (CZB) medium in the presence of 5 μM latrunculin A supplemented with trichostatin A (50 nM) for 9 hours (trichostatin A was not used however for the establishment of ntES cells from MEF3). After three washes in CZB, cloned embryos were cultured for 4 days in the same medium. Upon development to the morula or blastocyst stages, these embryos were used to establish ntES cell lines as described previously 21 with a slight modification in which 20% KnockOut Serum Replacement (Life Technologies, Rockville, MD, http://www.lifetech.com) plus 0.1 mg/mL adrenocorticotropic hormone fragments (1–24) (American Peptide Company, Sunnyvale, CA) was added to the ES cell medium instead of fetal calf serum 22.
Whole Genome Sequencing
Genomic DNA was prepared from established iPSCs, ntESCs, and sublines within passage 6 using DNeasy (Qiagen, Hilden, Germany, http://www1.qiagen.com). Whole genome sequencing (WGS) and subsequent computational analysis were performed as previously described 14. Illumina TruSeq DNA prep kits were used for library construction according to the protocol recommended by the manufacturer. Sequencing was performed using a HiSeq2000 sequencer (Illumina, San Diego, CA, https://www.illumina.com/) with 101‐base, paired‐end reads that were mapped to the C57BL/6 genome (NCBI37/mm9) with the Burrows–Wheeler Alignment tool (version 0.5.9) 23. We allowed up to two base mismatches in the mapping and uniquely mapped reads used for the subsequent analysis. Single nucleotide variation (SNV) candidates were identified using the CLC Genomics Workbench (Qiagen). The parameters were set as follows: minimum quality of central base, 30; minimum average quality of surrounding bases, 15; window length, 11; maximum gap and mismatch count, 2. Known single nucleotide polymorphisms (dbSNP128) and common SNVs among the parental cells were excluded from the panel of candidates. We also removed common SNVs among sister clones and false‐positives such as SNV candidates located within repetitive sequences or generated from sequencing errors. Our data, including bioinformatic results, were verified by Sanger sequencing of 106 SNV candidates 14. It has been demonstrated previously that this level of read depth is sufficient for genome‐wide SNV identification from inbred mouse genomes 13, 14, 24. We carried out WGS (mean mapped‐read depths of 11.8–15.1 for ntESCs and 21.8–25.3 for parental somatic cells) and called SNVs with a depth of at least 12–15 (Table 1). The three ntESC lines (B6mt‐1, ‐2, ‐3) were independently generated from identical TTF prepared from a single mouse. Because we generated these three lines in the initial stage of our study, genomic DNA was unavailable from the parent fibroblasts and we called the SNVs according to the methods used for other ntESCs and iPSCs with minor modifications, that is, complete removal of common SNVs but without comparison with their parent fibroblast genomes. This SNV‐calling procedure was confirmed by control lines.
| Cells | Total reads | Minimum depth | Shared genome coverage with parental somatic cells (%) | SNVs | SNVs/109 bp genome |
|---|---|---|---|---|---|
| MEF3 (parent) | 694,483,964 | ||||
| B6M3‐ntES‐1 | 375,292,092 | 12 | 57 | 310 | 211.7 |
| B6M3‐ntES‐2 | 401,137,062 | 13 | 55 | 274 | 194.6 |
| MEF5 (parent) | 761,283,214 | ||||
| e8M5‐ntES‐1 | 393,108,820 | 12 | 58 | 367 | 245.9 |
| e8M5‐ntES‐2 | 377,005,562 | 12 | 56 | 243 | 170.5 |
| e8M5‐ntES‐3 | 459,287,500 | 14 | 56 | 182 | 126.6 |
| TT3 (parent) | 737,450,498 | ||||
| e7‐ntES3‐2 | 479,868,420 | 15 | 57 | 241 | 165.8 |
| e7‐ntES3‐5 | 484,150,652 | 15 | 57 | 112 | 76.9 |
| e7‐ntES3‐6 | 458,349,670 | 14 | 57 | 87 | 60.0 |
| TT4 (parent) | |||||
| B6mt‐1 | 442,729,646 | 13 | 39 | 81 | 82.2 |
| B6mt‐2 | 445,800,148 | 13 | 39 | 86 | 87.3 |
| B6mt‐3 | 417,183,402 | 13 | 39 | 129 | 131.0 |
- Abbreviations: MEF, mouse embryonic fibroblast; ntES, nuclear transfer embryonic stem cell; SNV, single nucleotide variation.
Variant Allele Frequency Analysis
We performed a deeper level of WGS using mean mapped‐read depths of more than 40 to identify lower frequency SNVs as well as approximately 50% SNVs. The Illumina TruSeq DNA PCR‐free kit was used to prepare the library for WGS, and sequencing was performed using a HiSeq X sequencer (Illumina) with 150‐base, paired‐end reads that were mapped to the C57BL/6 genome (NCBI37/mm10) with the Isaac aligner (version 01.15.02.08). The mean mapped‐read depth for e6‐ntES2‐2 cells was 43 and that for the e6‐ntES2‐3 line was 42. Then SNVs were called using the Isaac variant caller (ver 2.0.13). A hand filter that was reported previously 14 was used to remove false‐positives, and SNVs were finally identified with the following conditions: minimum read depth of the variant sites, 20 and variant allele frequency (VAF), 10%–65%.
For subsequent amplicon sequencing, the genomic region containing the target SNV site was amplified by polymerase chain reaction (PCR) in a reaction containing 15 ng and 30 ng of genomic DNA prepared from ntES cells (approximately 2,500 cells) and parental somatic cells (approximately 5,000 cells), respectively. PCR was performed using Titanium‐taq DNA polymerase (Clontech, Palo Alto, CA, http://www.clontech.com), and the PCR conditions were as follows: 95°C for 1 minute, and then 32 cycles of 95°C for 20 seconds, 68°C for 30 seconds, and 72°C for 30 seconds. The PCR products were mixed and purified using a MinElute PCR Purification Kit (Qiagen). These amplicons were then subjected to comprehensive sequencing using a HiSeq 2000 sequencer (Illumina).
Results
Point Mutations in ntESCs Derived from Mouse Embryonic Fibroblasts
To better identify the point mutations that arise during iPSC generation, we previously improved the analysis system to more effectively identify de novo point mutations. Using this approach, we found that the ES cell genome has a few point mutations only, indicating that our system efficiently excludes pre‐existing SNVs as point mutation candidates 14.
In our present analysis, we examined another type of pluripotent stem cell generated by somatic cell nuclear transfer (SCNT), ntES cells (Supporting Information Fig. S1) 21. Using MEFs, 21 ntES cell lines were successfully generated and 5 were analyzed at random using WGS (Fig. 1A). These five cell lines were derived from different MEF fractions (MEF3 and 5) each of which was prepared from an independent single embryo. The B6M3ntES‐1 and ‐2 lines were derived from MEF3 and e8M5ntES‐1, ‐2 and ‐3 from MEF5 (Table 1). For the WGS analysis, we prepared the genomes as soon as possible after the colony pick. Each colony was expanded for three passages and then collected for genome preparation.
Point mutations revealed by WGS of ntESCs derived from MEFs. (A): Experimental design for the mutational analysis. (B): SNVs in each chromosome are shown, and the total number of SNVs is indicated in the parentheses. (C): Point mutation frequencies identified in ntESCs derived from MEFs. Error bars, SD; *, p < .05. The Mann–Whitney U test was used for the statistical analysis. (D) Base substitution profiles. Abbreviations: iPSCs, induced pluripotent stem cells; MEF, mouse embryonic fibroblast; ntESCs, nuclear transfer embryonic stem cells; WGS, whole genome sequencing; SNVs, single nucleotide variation; Ts, transition; Tv, transversion.
A direct comparison between each ntESC line and its corresponding parental MEFs revealed a significant number of point mutations in each genome: 310 SNVs in B6M3ntES‐1, 274 in B6M3ntES‐2, 367 in e8M5ntES‐1, 243 in e8M5ntES‐2, and 182 in e8M5ntES‐3 (Fig. 1B; Table 1). Interestingly, the number of point mutations, at 189.9 ± 44.8 SNVs/109 bp (Fig. 1C), was lower than those observed in iPS cells we generated previously using a retroviral mediated gene transduction method (329.7 ± 75.0 SNVs/109 bp) 14. Notably, transversion‐predominant base substitutions were evident in ntESCs, as had been observed previously in iPSCs (Fig. 1D) 14.
Point Mutations in the ntESCs Generated from Tail Fibroblasts
To examine whether the parental somatic cell type influences the point mutation frequency, we also generated ntESCs from TTFs and again analyzed the genomes via WGS (Fig. 2A). Surprisingly, we observed only 81–241 SNVs in these TTF‐derived ntESC genomes (Fig. 2B). This was approximately 50% lower than the number of SNVs observed in MEF‐derived ntESCs (100.5 ± 39.7 SNVs/109 bp; Fig. 2C). Furthermore, two ntESC lines, e7‐ntES3‐6 and e7‐ntES3‐5, showed only 60.0 and 76.9 SNVs/109 bp, respectively, which was approximately 80% less than the number observed in the iPSCs that we previously generated using retroviral mediated gene delivery (329.7 ± 75.0 SNVs/109 bp) 14.
Point mutations in ntESCs derived from mouse TTFs. (A): Experimental design. (B): SNVs in each chromosome are shown, and the total number of SNVs is indicated in the parentheses. (C): Point mutation frequencies identified in ntESCs derived from MEFs, TTFs, and ESCs. Point mutations in ESCs were previously reported 14. Error bars, SD; *, p < .05. The Mann–Whitney U test was used for the statistical analysis. (D): Base substitution profiles. Abbreviations: ntESCs, nuclear transfer embryonic stem cells; MEF, mouse embryonic fibroblasts; SNVs, single nucleotide variation; TTF, tail‐tip fibroblasts; Ts, transition; Tv, transversion; WGS, whole genome sequencing.
Base substitution profile analysis of these clones also clearly indicated a reduced point mutation frequency, moving from a transversion‐predominance to a transition‐predominance (Figs. 1D, 2D). Notably, the B6mt‐3 clone showed a transition‐predominance in SNVs, similar to ES cells. This is the first example of a clone that we generated from a genome reprogrammed cell line exhibiting transition‐predominant point mutations. Thus, we found when using mouse tail fibroblasts that only a small number of point mutations arise during genome reprogramming via nuclear transfer. This strongly suggested that the parental cell type affects the point mutation frequency.
On the other hand, iPSCs were also generated from the somatic cells that were used for our current ntESC generation by retroviral gene transduction (Supporting Information Fig. S2). We compared the point mutation frequencies in the ntES cells with those in other types of pluripotent stem cells that were reported previously, that is, retrovirus‐mediated iPSCs, integration‐free iPSCs, and ES cells (Supporting Information Fig. S3) 14. All of the cell lines in this comparison were generated using the same mouse strain, C57BL/6J. Thus, our present findings indicate that the number of point mutations in iPSCs and ntESCs depends on the procedure and somatic cell type employed for their generation.
No Protein‐Coding Region Bias
We identified a total of 2,112 SNVs by WGS of 11 ntESC lines (Supporting Information Table S1). There was little or no bias with regard to the coding regions (exons) (Supporting Information Fig. S4). Only 25 of these SNVs (17 nonsynonymous and 8 synonymous) were identified in exons, 1.19% of the total panel (Table 2). Furthermore, no oncogenes or anti‐oncogenes were found among the genes in which these 25 SNVs were identified (Cosmic Cancer gene census data) 25. Moreover, it is noteworthy that no SNVs were observed in the coding regions in 4 out of the 11 ntESC lines (B6M3‐ntES‐1, e8M5‐ntES‐1, B6mt‐1, and B6mt‐2).
| Cell | Chromosome | Position | Wild allele (+ strand) | Mutant allele (+ strand) | Gene | Strand of transcript | Codon (amino acid) | SNV type |
|---|---|---|---|---|---|---|---|---|
| B6M3‐ntES‐1 | Not identified | |||||||
| B6M3‐ntES‐2 | chr7 | 111,737,220 | C | A | Olfr654 | + | CGG(R)‐CGt(R) | Synonymous |
| chr9 | 109,386,923 | G | T | Fbxw19 | – | CAC(H)‐CAa(Q) | Nonsynonymous | |
| chr16 | 38,450,336 | G | A | Afprh | – | ACC(T)‐ACt(T) | Synonymous | |
| chr17 | 57,684,729 | A | C | Vmn2r120 | – | ATT(I)‐ATg(M) | Nonsynonymous | |
| e8M5‐ntES‐1 | Not identified | |||||||
| e8M5‐ntES‐2 | chr1 | 66,726,256 | C | T | Unc80 | + | AGC(S)‐AGt(S) | Synonymous |
| chr1 | 120,763,248 | G | A | Gli2 | – | CTC(L)‐tTC(F) | Nonsynonymous | |
| chr2 | 91,650,252 | G | A | Ambra1 | + | CGC(R)‐CaC(H) | Nonsynonymous | |
| chr15 | 82,685,521 | T | G | Tcf20 | – | AGT(S)‐cGT(R) | Nonsynonymous | |
| e8M5‐ntES‐3 | chr2 | 92,299,226 | C | G | Slc35c1 | – | GTC(V)‐cTC(L) | Nonsynonymous |
| chr5 | 147,395,063 | G | A | Gpr12 | – | TTC(F)‐TTt(F) | Synonymous | |
| chr13 | 22,074,919 | T | C | Pom121l2 | + | GTT(V)‐GcT(A) | Nonsynonymous | |
| chr14 | 35,392,456 | C | G | Ldb3 | – | GCC(A)‐cCC(P) | Nonsynonymous | |
| e7‐ntES3‐2 | chr4 | 57,081,192 | T | A | Epb4.1l4b | – | GAA(E)‐GAt(D) | Nonsynonymous |
| chr10 | 76,172,669 | C | T | Col6a1 | – | GAT(D)‐aAT(N) | Nonsynonymous | |
| chr15 | 82,235,997 | A | G | Cyp2d10 | – | TTT(F)‐cTT(L) | Nonsynonymous | |
| chr19 | 53,888,632 | G | A | Rbm20 | + | AGG(R)‐AGa(R) | Synonymous | |
| e7‐ntES3‐5 | chr4 | 134,323,782 | T | A | Ldlrap1 | – | ATC(I)‐tTC(F) | Nonsynonymous |
| chr7 | 6,135,774 | G | A | Zfp444 | + | GAG(E)‐GAa(E) | Nonsynonymous | |
| chr10 | 127,556,534 | A | G | Baz2a | + | GAT(D)‐GgT(G) | Nonsynonymous | |
| chr17 | 25,243,828 | G | A | Telo2 | – | GGC(G)‐GGt(G) | Synonymous | |
| e7‐ntES3‐6 | chr7 | 18,765,148 | C | A | Igfl3 | + | CGA(R)‐aGA(R) | Synonymous |
| chr7 | 48,544,101 | C | G | Gm5592 | + | GCT(A)‐GgT(G) | Nonsynonymous | |
| B6mt‐1 | Not identified | |||||||
| B6mt‐2 | Not identified | |||||||
| B6mt‐3 | chr9 | 106,335,483 | G | A | Acy1 | – | CGT(R)‐tGT(C) | Nonsynonymous |
| chr12 | 114,217,909 | G | A | Brf1 | – | AAC(N)‐AAt(N) | Synonymous | |
| chr16 | 58,926,151 | A | T | Olfr181 | – | CTA(L)‐CAA(Q) | Nonsynonymous | |
- Abbreviations: ntES, nuclear transfer embryonic stem cell; SNV, single nucleotide variation.
Variant Allele Frequency of Point Mutations in ntES Cells
We investigated the VAF of SNVs identified in our ntES cells. A deeper WGS approach and subsequent amplicon sequencing analysis with next generation DNA sequencing system were required in this case because it is not feasible to identify SNVs with a VAF below 50% using standard WGS. We conducted VAF analysis on three ntES cells, B6M3‐ntES‐1 derived from MEFs, and e6‐ntES2‐2 and e6‐ntES2‐3 derived from TTFs. We additionally performed WGS of the e6‐ntES2‐2 and e6‐ntES2‐3 lines. We generated a VAF histogram for the SNV candidates we detected and small but clear peaks were evident at the 25%–30% frequency in all three lines in addition to the main peak at the approximately 50% allele frequency (Fig. 3A). Notably, more than 40 mean mapped‐read depths was required for detecting the approximately 25% allele frequency peak in the e6‐ntES2‐2 and e6‐ntES2‐3 clones derived from TTFs.
Allele frequencies of SNVs. (A): VAF values suggested by WGS results. The ratios of transversion to transition for each peak are also shown (above). (B): VAF values indicated by amplicon sequencing. This analysis focused on the shoulder peaks indicated from the WGS data. (C): No pre‐existing SNVs were detectable in the parental somatic cell genomes. Abbreviations: ntESCs, nuclear transfer embryonic SNVs, single nucleotide variation; VAF, variant allele frequency; WGS, whole genome sequencing.
Subsequently, we conducted amplicon sequencing of randomly chosen SNVs covering less than the 50% range. These included 18 SNVs for B6M3‐ntES‐1, 42 SNVs for e6‐ntES2‐2, and 32 SNVs for e6‐ntES2‐3. We then confirmed the presence of significant numbers of SNVs, with a VAF below 50% (Fig. 3B), in their genomes. It is notable that a VAF of less than 50% is often affected by the cell growth rate within an ntESC colony in addition to the PCR efficiency, which can be affected by the DNA sequences encompassing SNV candidates. On the other hand, although we also examined the VAF of these SNVs in their parent somatic cell genomes using amplicon sequencing, no signal was detected, indicating that these mutations were not pre‐existing (Fig. 3C). Importantly, a transversion‐predominance was evident in the less than 50% SNVs in all ntESC lines examined.
In summary therefore, our finding of the existence of low‐frequency SNVs in ntES cells, of the lack of SNV candidates in the parent somatic cell genomes, and our detection of unique base substitutions in a transversion‐predominant fashion indicate that these SNVs are de novo and are associated with ntES cell generation.
Discussion
We have expanded on the previous observations that somatic variants appear in iPSCs to reveal that they are also present in ESCs generated by SCNT. Our findings thus illustrate that this is a phenomenon intrinsic to cellular reprogramming and not simply a secondary effect of reprogramming with Yamanaka factors. Similar results from the exome sequencing of ntESC genomes were also recently reported 26, 27. Point mutation studies of iPSCs can therefore directly contribute to our understanding of the molecular basis of genome reprogramming.
A significant number of SNVs were identified in each of our current reprogrammed pluripotent stem cell lines. However, each of our lines showed a unique SNV profile with the exception of 98 SNVs identified in the B6M3ntES‐1 and B6M3ntES‐2 sister clones. However, we identified these SNVs also just before nuclear transfer in the parent cells, which were obtained after several passages of the MEFs used as WGS templates, by random sampling testing with amplicon sequencing (data not shown). Notably in this case, we passaged the parental cells for longer than usual because this was our first attempt to generate ntES cells from MEFs. Other than this experiment, we have never identified common SNVs in stem cells generated using MEFs or TTFs. In this study, we identified 2,112 noncommon SNVs in total from 11 reprogrammed pluripotent stem cell lines.
A unique, transversion‐predominant base substitution profile has been observed in all but one, B6mt‐3, of the stem cell lines generated from MEFs or TTFs using various methods in our current and previous studies. We therefore performed bioinformatic analysis on the WGS results from iPSCs generated by another group from MEFs or adipocyte progenitor cells (APCs) using Tet‐on inducible reprogramming 15. Significantly, this independent analysis identified 954 and 1,212 of SNVs in iPSC lines derived from APCs (2°‐iPSC‐32) and MEFs (5°‐iPSC‐8), respectively, that showed a definitive transversion‐predominance in both cases (Supporting Information Fig. S5).
On the other hand, the issue of whether the frequency of point mutations in iPSCs and ntESCs can be reduced has remained an important concern, since these mutations are potentially problematic for the use of such stem cells in regenerative medicine. Our current findings suggest however that we can significantly reduce the point mutation frequency in induced pluripotent stem cells by optimizing certain experimental conditions such as the gene delivery method and choice of parent somatic cell type. It was a particularly interesting and notable finding in this regard that stem cell lines in which only a small number of point mutations arose could be established from mouse tail fibroblasts using nuclear transfer genome reprogramming. Considering that our TTFs were prepared from an adult mouse which would necessarily contain more pre‐existing SNVs than MEFs derived from an E13.5 embryo, the choice of parent cell seems to be more crucial for controlling the frequency of point mutations.
Taken together, our present results suggest the possibility that the number of point mutations may be reduced in reprogrammed pluripotent stem cells. In addition, most of the point mutations in such stem cells seem not to be essential for genome reprogramming, but to simply be by‐products of this process. A better understanding of the molecular mechanisms underlying the occurrence of point mutations in iPSCs and ntESCs will be essential to possibly developing point mutation‐free, and hence clinically ideal, pluripotent stem cells in the future.
Conclusion
A significant number of point mutations have been reported in all iPSC genome studies to date which has raised considerable concerns over the safety of the use of these stem cells in a clinical application. Reducing this point mutation frequency will be crucial if these cells are to be used in regenerative medicine in the future. Significantly, ES cells generated by SCNT from tail fibroblasts have far fewer point mutations than other iPSCs. A considerable reduction in the point mutation frequency in these stem cells may therefore be feasible by optimizing the genome reprogramming conditions.
Acknowledgments
We thank H. Hirakawa for help with the teratoma assay and Y. Une (Azabu University) for assistance with the histological diagnosis; M. Uda, H. Yoshida, K. Nishikawa, A. Ishibashi, and Y. Shindo for technical assistance. This work was partially funded by the Japan Society for the Promotion of Science (Grant Number KAKENHI 25290068, 21688020, and 23248048), the Takeda Science Foundation, Uehara Memorial Foundation, and Mitsubishi Foundation.
Author Contributions
R.A. and M.A.: conception and design, financial support, data analysis and interpretation, manuscript writing, and final approval of manuscript; E.M., Y.H., S.W., and H.N.: collection and assembly of data; M.S. and M.N.: collection of data; Y.K.: data analysis and interpretation; T.W.: collection and assembly of data, financial support, and data analysis and interpretation.
Accession Numbers
Raw Illumina sequencing reads are deposited at the DDBJ Sequence Read Archive (accession numbers DRA002956, DRA003544, DRA005297, and DRA005423).
Disclosure of Potential Conflicts of Interest
The authors indicate no potential conflicts of interest.
